Abstract
Deep learning models such as convolutional neural networks and transformers have been widely applied to solve 3D object detection problems in the domain of autonomous driving. While existing models have achieved outstanding performance on most open benchmarks, the generalization ability of these deep networks is still in doubt. To adapt models to other domains including different cities, countries, and weather, retraining with the target domain data is currently necessary, which hinders the wide application of autonomous driving. In this paper, we deeply analyze the cross-domain performance of the state-of-the-art models. We observe that most models will overfit the training domains and it is challenging to adapt them to other domains directly. Existing domain adaptation methods for 3D object detection problems are actually shifting the models' knowledge domain instead of improving their generalization ability. We then propose additional evaluation metrics -- the side-view and front-view AP -- to better analyze the core issues of the methods' heavy drops in accuracy levels. By using the proposed metrics and further evaluating the cross-domain performance in each dimension, we conclude that the overfitting problem happens more obviously on the front-view surface and the width dimension which usually faces the sensor and has more 3D points surrounding it. Meanwhile, our experiments indicate that the density of the point cloud data also significantly influences the models' cross-domain performance.
Title:
CatFree3D: Category-agnostic 3D Object Detection with Diffusion
Authors: Wenjing Bian, Zirui Wang, Andrea Vedaldi
Subjects: Subjects:
Computer Vision and Pattern Recognition (cs.CV)
Abstract
Image-based 3D object detection is widely employed in applications such as autonomous vehicles and robotics, yet current systems struggle with generalisation due to complex problem setup and limited training data. We introduce a novel pipeline that decouples 3D detection from 2D detection and depth prediction, using a diffusion-based approach to improve accuracy and support category-agnostic detection. Additionally, we introduce the Normalised Hungarian Distance (NHD) metric for an accurate evaluation of 3D detection results, addressing the limitations of traditional IoU and GIoU metrics. Experimental results demonstrate that our method achieves state-of-the-art accuracy and strong generalisation across various object categories and datasets.
Title:
Symmetric masking strategy enhances the performance of Masked Image Modeling
Abstract
Masked Image Modeling (MIM) is a technique in self-supervised learning that focuses on acquiring detailed visual representations from unlabeled images by estimating the missing pixels in randomly masked sections. It has proven to be a powerful tool for the preliminary training of Vision Transformers (ViTs), yielding impressive results across various tasks. Nevertheless, most MIM methods heavily depend on the random masking strategy to formulate the pretext task. This strategy necessitates numerous trials to ascertain the optimal dropping ratio, which can be resource-intensive, requiring the model to be pre-trained for anywhere between 800 to 1600 epochs. Furthermore, this approach may not be suitable for all datasets. In this work, we propose a new masking strategy that effectively helps the model capture global and local features. Based on this masking strategy, SymMIM, our proposed training pipeline for MIM is introduced. SymMIM achieves a new SOTA accuracy of 85.9\% on ImageNet using ViT-Large and surpasses previous SOTA across downstream tasks such as image classification, semantic segmentation, object detection, instance segmentation tasks, and so on.
Title:
DeTPP: Leveraging Object Detection for Robust Long-Horizon Event Prediction
Abstract
Forecasting future events over extended periods, known as long-horizon prediction, is a fundamental task in various domains, including retail, finance, healthcare, and social networks. Traditional methods, such as Marked Temporal Point Processes (MTPP), typically use autoregressive models to predict multiple future events. However, these models frequently encounter issues such as converging to constant or repetitive outputs, which significantly limits their effectiveness and applicability. To overcome these limitations, we propose DeTPP (Detection-based Temporal Point Processes), a novel approach inspired by object detection methods from computer vision. DeTPP utilizes a novel matching-based loss function that selectively focuses on reliably predictable events, enhancing both training robustness and inference diversity. Our method sets a new state-of-the-art in long-horizon event prediction, significantly outperforming existing MTPP and next-K approaches. The implementation of DeTPP is publicly available on GitHub.
Title:
MCTR: Multi Camera Tracking Transformer
Authors: Alexandru Niculescu-Mizil, Deep Patel, Iain Melvin
Subjects: Subjects:
Computer Vision and Pattern Recognition (cs.CV)
Abstract
Multi-camera tracking plays a pivotal role in various real-world applications. While end-to-end methods have gained significant interest in single-camera tracking, multi-camera tracking remains predominantly reliant on heuristic techniques. In response to this gap, this paper introduces Multi-Camera Tracking tRansformer (MCTR), a novel end-to-end approach tailored for multi-object detection and tracking across multiple cameras with overlapping fields of view. MCTR leverages end-to-end detectors like DEtector TRansformer (DETR) to produce detections and detection embeddings independently for each camera view. The framework maintains set of track embeddings that encaplusate global information about the tracked objects, and updates them at every frame by integrating the local information from the view-specific detection embeddings. The track embeddings are probabilistically associated with detections in every camera view and frame to generate consistent object tracks. The soft probabilistic association facilitates the design of differentiable losses that enable end-to-end training of the entire system. To validate our approach, we conduct experiments on MMPTrack and AI City Challenge, two recently introduced large-scale multi-camera multi-object tracking datasets.
Keyword: transformer
Title:
AI-driven Transformer Model for Fault Prediction in Non-Linear Dynamic Automotive System
Abstract
Fault detection in automotive engine systems is one of the most promising research areas. Several works have been done in the field of model-based fault diagnosis. Many researchers have discovered more advanced statistical methods and algorithms for better fault detection on any automotive dynamic engine system. The gas turbines/diesel engines produce highly complex and huge data which are highly non-linear. So, researchers should come up with an automated system that is more resilient and robust enough to handle this huge, complex data in highly non-linear dynamic automotive systems. Here, I present an AI-based fault classification and prediction model in the diesel engine that can be applied to any highly non-linear dynamic automotive system. The main contribution of this paper is the AI-based Transformer fault classification and prediction model in the diesel engine concerning the worldwide harmonic light vehicle test procedure (WLTP) driving cycle. This model used 27 input dimensions, 64 hidden dimensions with 2 layers, and 9 heads to create a classifier with 12 output heads (one for fault-free data and 11 different fault types). This model was trained on the UTSA Arc High-Performance Compute (HPC) cluster with 5 NVIDIA V100 GPUs, 40-core CPUs, and 384GB RAM and achieved 70.01 % accuracy on a held test set.
Title:
Revisiting Cross-Domain Problem for LiDAR-based 3D Object Detection
Authors: Ruixiao Zhang, Juheon Lee, Xiaohao Cai, Adam Prugel-Bennett
Subjects: Subjects:
Computer Vision and Pattern Recognition (cs.CV)
Abstract
Deep learning models such as convolutional neural networks and transformers have been widely applied to solve 3D object detection problems in the domain of autonomous driving. While existing models have achieved outstanding performance on most open benchmarks, the generalization ability of these deep networks is still in doubt. To adapt models to other domains including different cities, countries, and weather, retraining with the target domain data is currently necessary, which hinders the wide application of autonomous driving. In this paper, we deeply analyze the cross-domain performance of the state-of-the-art models. We observe that most models will overfit the training domains and it is challenging to adapt them to other domains directly. Existing domain adaptation methods for 3D object detection problems are actually shifting the models' knowledge domain instead of improving their generalization ability. We then propose additional evaluation metrics -- the side-view and front-view AP -- to better analyze the core issues of the methods' heavy drops in accuracy levels. By using the proposed metrics and further evaluating the cross-domain performance in each dimension, we conclude that the overfitting problem happens more obviously on the front-view surface and the width dimension which usually faces the sensor and has more 3D points surrounding it. Meanwhile, our experiments indicate that the density of the point cloud data also significantly influences the models' cross-domain performance.
Title:
BankTweak: Adversarial Attack against Multi-Object Trackers by Manipulating Feature Banks
Abstract
Multi-object tracking (MOT) aims to construct moving trajectories for objects, and modern multi-object trackers mainly utilize the tracking-by-detection methodology. Initial approaches to MOT attacks primarily aimed to degrade the detection quality of the frames under attack, thereby reducing accuracy only in those specific frames, highlighting a lack of \textit{efficiency}. To improve efficiency, recent advancements manipulate object positions to cause persistent identity (ID) switches during the association phase, even after the attack ends within a few frames. However, these position-manipulating attacks have inherent limitations, as they can be easily counteracted by adjusting distance-related parameters in the association phase, revealing a lack of \textit{robustness}. In this paper, we present \textsf{BankTweak}, a novel adversarial attack designed for MOT trackers, which features efficiency and robustness. \textsf{BankTweak} focuses on the feature extractor in the association phase and reveals vulnerability in the Hungarian matching method used by feature-based MOT systems. Exploiting the vulnerability, \textsf{BankTweak} induces persistent ID switches (addressing \textit{efficiency}) even after the attack ends by strategically injecting altered features into the feature banks without modifying object positions (addressing \textit{robustness}). To demonstrate the applicability, we apply \textsf{BankTweak} to three multi-object trackers (DeepSORT, StrongSORT, and MOTDT) with one-stage, two-stage, anchor-free, and transformer detectors. Extensive experiments on the MOT17 and MOT20 datasets show that our method substantially surpasses existing attacks, exposing the vulnerability of the tracking-by-detection framework to \textsf{BankTweak}.
Title:
TReX- Reusing Vision Transformer's Attention for Efficient Xbar-based Computing
Authors: Abhishek Moitra, Abhiroop Bhattacharjee, Youngeun Kim, Priyadarshini Panda
Abstract
Due to the high computation overhead of Vision Transformers (ViTs), In-memory Computing architectures are being researched towards energy-efficient deployment in edge-computing scenarios. Prior works have proposed efficient algorithm-hardware co-design and IMC-architectural improvements to improve the energy-efficiency of IMC-implemented ViTs. However, all prior works have neglected the overhead and co-depencence of attention blocks on the accuracy-energy-delay-area of IMC-implemented ViTs. To this end, we propose TReX- an attention-reuse-driven ViT optimization framework that effectively performs attention reuse in ViT models to achieve optimal accuracy-energy-delay-area tradeoffs. TReX optimally chooses the transformer encoders for attention reuse to achieve near iso-accuracy performance while meeting the user-specified delay requirement. Based on our analysis on the Imagenet-1k dataset, we find that TReX achieves 2.3x (2.19x) EDAP reduction and 1.86x (1.79x) TOPS/mm2 improvement with ~1% accuracy drop in case of DeiT-S (LV-ViT-S) ViT models. Additionally, TReX achieves high accuracy at high EDAP reduction compared to state-of-the-art token pruning and weight sharing approaches. On NLP tasks such as CoLA, TReX leads to 2% higher non-ideal accuracy compared to baseline at 1.6x lower EDAP.
Title:
Symmetric masking strategy enhances the performance of Masked Image Modeling
Abstract
Masked Image Modeling (MIM) is a technique in self-supervised learning that focuses on acquiring detailed visual representations from unlabeled images by estimating the missing pixels in randomly masked sections. It has proven to be a powerful tool for the preliminary training of Vision Transformers (ViTs), yielding impressive results across various tasks. Nevertheless, most MIM methods heavily depend on the random masking strategy to formulate the pretext task. This strategy necessitates numerous trials to ascertain the optimal dropping ratio, which can be resource-intensive, requiring the model to be pre-trained for anywhere between 800 to 1600 epochs. Furthermore, this approach may not be suitable for all datasets. In this work, we propose a new masking strategy that effectively helps the model capture global and local features. Based on this masking strategy, SymMIM, our proposed training pipeline for MIM is introduced. SymMIM achieves a new SOTA accuracy of 85.9\% on ImageNet using ViT-Large and surpasses previous SOTA across downstream tasks such as image classification, semantic segmentation, object detection, instance segmentation tasks, and so on.
Title:
Open-Set Deepfake Detection: A Parameter-Efficient Adaptation Method with Forgery Style Mixture
Authors: Chenqi Kong, Anwei Luo, Peijun Bao, Haoliang Li, Renjie Wan, Zengwei Zheng, Anderson Rocha, Alex C. Kot
Subjects: Subjects:
Computer Vision and Pattern Recognition (cs.CV)
Abstract
Open-set face forgery detection poses significant security threats and presents substantial challenges for existing detection models. These detectors primarily have two limitations: they cannot generalize across unknown forgery domains and inefficiently adapt to new data. To address these issues, we introduce an approach that is both general and parameter-efficient for face forgery detection. It builds on the assumption that different forgery source domains exhibit distinct style statistics. Previous methods typically require fully fine-tuning pre-trained networks, consuming substantial time and computational resources. In turn, we design a forgery-style mixture formulation that augments the diversity of forgery source domains, enhancing the model's generalizability across unseen domains. Drawing on recent advancements in vision transformers (ViT) for face forgery detection, we develop a parameter-efficient ViT-based detection model that includes lightweight forgery feature extraction modules and enables the model to extract global and local forgery clues simultaneously. We only optimize the inserted lightweight modules during training, maintaining the original ViT structure with its pre-trained ImageNet weights. This training strategy effectively preserves the informative pre-trained knowledge while flexibly adapting the model to the task of Deepfake detection. Extensive experimental results demonstrate that the designed model achieves state-of-the-art generalizability with significantly reduced trainable parameters, representing an important step toward open-set Deepfake detection in the wild.
Title:
LIMP: Large Language Model Enhanced Intent-aware Mobility Prediction
Abstract
Human mobility prediction is essential for applications like urban planning and transportation management, yet it remains challenging due to the complex, often implicit, intentions behind human behavior. Existing models predominantly focus on spatiotemporal patterns, paying less attention to the underlying intentions that govern movements. Recent advancements in large language models (LLMs) offer a promising alternative research angle for integrating commonsense reasoning into mobility prediction. However, it is a non-trivial problem because LLMs are not natively built for mobility intention inference, and they also face scalability issues and integration difficulties with spatiotemporal models. To address these challenges, we propose a novel LIMP (LLMs for Intent-ware Mobility Prediction) framework. Specifically, LIMP introduces an "Analyze-Abstract-Infer" (A2I) agentic workflow to unleash LLM's commonsense reasoning power for mobility intention inference. Besides, we design an efficient fine-tuning scheme to transfer reasoning power from commercial LLM to smaller-scale, open-source language model, ensuring LIMP's scalability to millions of mobility records. Moreover, we propose a transformer-based intention-aware mobility prediction model to effectively harness the intention inference ability of LLM. Evaluated on two real-world datasets, LIMP significantly outperforms baseline models, demonstrating improved accuracy in next-location prediction and effective intention inference. The interpretability of intention-aware mobility prediction highlights our LIMP framework's potential for real-world applications. Codes and data can be found in this https URL .
Title:
IAA: Inner-Adaptor Architecture Empowers Frozen Large Language Model with Multimodal Capabilities
Authors: Bin Wang, Chunyu Xie, Dawei Leng, Yuhui Yin
Subjects: Subjects:
Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Machine Learning (cs.LG)
Abstract
In the field of multimodal large language models (MLLMs), common methods typically involve unfreezing the language model during training to foster profound visual understanding. However, the fine-tuning of such models with vision-language data often leads to a diminution of their natural language processing (NLP) capabilities. To avoid this performance degradation, a straightforward solution is to freeze the language model while developing multimodal competencies. Unfortunately, previous works have not attained satisfactory outcomes. Building on the strategy of freezing the language model, we conduct thorough structural exploration and introduce the Inner-Adaptor Architecture (IAA). Specifically, the architecture incorporates multiple multimodal adaptors at varying depths within the large language model to facilitate direct interaction with the inherently text-oriented transformer layers, thereby enabling the frozen language model to acquire multimodal capabilities. Unlike previous approaches of freezing language models that require large-scale aligned data, our proposed architecture is able to achieve superior performance on small-scale datasets. We conduct extensive experiments to improve the general multimodal capabilities and visual grounding abilities of the MLLM. Our approach remarkably outperforms previous state-of-the-art methods across various vision-language benchmarks without sacrificing performance on NLP tasks. Code and models are available at this https URL.
Title:
State-of-the-Art Fails in the Art of Damage Detection
Authors: Daniela Ivanova, Marco Aversa, Paul Henderson, John Williamson
Subjects: Subjects:
Computer Vision and Pattern Recognition (cs.CV)
Abstract
Accurately detecting and classifying damage in analogue media such as paintings, photographs, textiles, mosaics, and frescoes is essential for cultural heritage preservation. While machine learning models excel in correcting global degradation if the damage operator is known a priori, we show that they fail to predict where the damage is even after supervised training; thus, reliable damage detection remains a challenge. We introduce DamBench, a dataset for damage detection in diverse analogue media, with over 11,000 annotations covering 15 damage types across various subjects and media. We evaluate CNN, Transformer, and text-guided diffusion segmentation models, revealing their limitations in generalising across media types.
Title:
Accuracy Improvement of Cell Image Segmentation Using Feedback Former
Authors: Hinako Mitsuoka, Kazuhiro Hotta
Subjects: Subjects:
Computer Vision and Pattern Recognition (cs.CV)
Abstract
Semantic segmentation of microscopy cell images by deep learning is a significant technique. We considered that the Transformers, which have recently outperformed CNNs in image recognition, could also be improved and developed for cell image segmentation. Transformers tend to focus more on contextual information than on detailed information. This tendency leads to a lack of detailed information for segmentation. Therefore, to supplement or reinforce the missing detailed information, we hypothesized that feedback processing in the human visual cortex should be effective. Our proposed Feedback Former is a novel architecture for semantic segmentation, in which Transformers is used as an encoder and has a feedback processing mechanism. Feature maps with detailed information are fed back to the lower layers from near the output of the model to compensate for the lack of detailed information which is the weakness of Transformers and improve the segmentation accuracy. By experiments on three cell image datasets, we confirmed that our method surpasses methods without feedback, demonstrating its superior accuracy in cell image segmentation. Our method achieved higher segmentation accuracy while consuming less computational cost than conventional feedback approaches. Moreover, our method offered superior precision without simply increasing the model size of Transformer encoder, demonstrating higher accuracy with lower computational cost.
Title:
Atlas Gaussians Diffusion for 3D Generation with Infinite Number of Points
Abstract
Using the latent diffusion model has proven effective in developing novel 3D generation techniques. To harness the latent diffusion model, a key challenge is designing a high-fidelity and efficient representation that links the latent space and the 3D space. In this paper, we introduce Atlas Gaussians, a novel representation for feed-forward native 3D generation. Atlas Gaussians represent a shape as the union of local patches, and each patch can decode 3D Gaussians. We parameterize a patch as a sequence of feature vectors and design a learnable function to decode 3D Gaussians from the feature vectors. In this process, we incorporate UV-based sampling, enabling the generation of a sufficiently large, and theoretically infinite, number of 3D Gaussian points. The large amount of 3D Gaussians enables high-quality details of generation results. Moreover, due to local awareness of the representation, the transformer-based decoding procedure operates on a patch level, ensuring efficiency. We train a variational autoencoder to learn the Atlas Gaussians representation, and then apply a latent diffusion model on its latent space for learning 3D Generation. Experiments show that our approach outperforms the prior arts of feed-forward native 3D generation.
Title:
Multivariate Time-Series Anomaly Detection based on Enhancing Graph Attention Networks with Topological Analysis
Abstract
Unsupervised anomaly detection in time series is essential in industrial applications, as it significantly reduces the need for manual intervention. Multivariate time series pose a complex challenge due to their feature and temporal dimensions. Traditional methods use Graph Neural Networks (GNNs) or Transformers to analyze spatial while RNNs to model temporal dependencies. These methods focus narrowly on one dimension or engage in coarse-grained feature extraction, which can be inadequate for large datasets characterized by intricate relationships and dynamic changes. This paper introduces a novel temporal model built on an enhanced Graph Attention Network (GAT) for multivariate time series anomaly detection called TopoGDN. Our model analyzes both time and feature dimensions from a fine-grained perspective. First, we introduce a multi-scale temporal convolution module to extract detailed temporal features. Additionally, we present an augmented GAT to manage complex inter-feature dependencies, which incorporates graph topology into node features across multiple scales, a versatile, plug-and-play enhancement that significantly boosts the performance of GAT. Our experimental results confirm that our approach surpasses the baseline models on four datasets, demonstrating its potential for widespread application in fields requiring robust anomaly detection. The code is available at this https URL.
Title:
NEST: Self-supervised Fast Conformer as All-purpose Seasoning to Speech Processing Tasks
Authors: He Huang, Taejin Park, Kunal Dhawan, Ivan Medennikov, Krishna C. Puvvada, Nithin Rao Koluguri, Weiqing Wang, Jagadeesh Balam, Boris Ginsburg
Subjects: Subjects:
Sound (cs.SD); Audio and Speech Processing (eess.AS)
Abstract
Self-supervised learning has been proved to benefit a wide range of speech processing tasks, such as speech recognition/translation, speaker verification and diarization, etc. However, most of these approaches are computationally intensive due to using transformer encoder and lack of sub-sampling. In this paper, we propose a new self-supervised learning model termed as Neural Encoder for Self-supervised Training (NEST). Specifically, we adopt the FastConformer architecture, which has an 8x sub-sampling rate and is faster than Transformer or Conformer architectures. Instead of clustering-based token generation, we resort to fixed random projection for its simplicity and effectiveness. We also propose a generalized noisy speech augmentation that teaches the model to disentangle the main speaker from noise or other speakers. Experiments show that the proposed NEST model improves over existing self-supervised models on a variety of speech processing tasks. Code and checkpoints will be publicly available via NVIDIA NeMo toolkit.
Title:
Long-Term Pre-training for Temporal Action Detection with Transformers
Authors: Jihwan Kim, Miso Lee, Jae-Pil Heo
Subjects: Subjects:
Computer Vision and Pattern Recognition (cs.CV)
Abstract
Temporal action detection (TAD) is challenging, yet fundamental for real-world video applications. Recently, DETR-based models for TAD have been prevailing thanks to their unique benefits. However, transformers demand a huge dataset, and unfortunately data scarcity in TAD causes a severe degeneration. In this paper, we identify two crucial problems from data scarcity: attention collapse and imbalanced performance. To this end, we propose a new pre-training strategy, Long-Term Pre-training (LTP), tailored for transformers. LTP has two main components: 1) class-wise synthesis, 2) long-term pretext tasks. Firstly, we synthesize long-form video features by merging video snippets of a target class and non-target classes. They are analogous to untrimmed data used in TAD, despite being created from trimmed data. In addition, we devise two types of long-term pretext tasks to learn long-term dependency. They impose long-term conditions such as finding second-to-fourth or short-duration actions. Our extensive experiments show state-of-the-art performances in DETR-based methods on ActivityNet-v1.3 and THUMOS14 by a large margin. Moreover, we demonstrate that LTP significantly relieves the data scarcity issues in TAD.
Title:
EAViT: External Attention Vision Transformer for Audio Classification
Abstract
This paper presents the External Attention Vision Transformer (EAViT) model, a novel approach designed to enhance audio classification accuracy. As digital audio resources proliferate, the demand for precise and efficient audio classification systems has intensified, driven by the need for improved recommendation systems and user personalization in various applications, including music streaming platforms and environmental sound recognition. Accurate audio classification is crucial for organizing vast audio libraries into coherent categories, enabling users to find and interact with their preferred audio content more effectively. In this study, we utilize the GTZAN dataset, which comprises 1,000 music excerpts spanning ten diverse genres. Each 30-second audio clip is segmented into 3-second excerpts to enhance dataset robustness and mitigate overfitting risks, allowing for more granular feature analysis. The EAViT model integrates multi-head external attention (MEA) mechanisms into the Vision Transformer (ViT) framework, effectively capturing long-range dependencies and potential correlations between samples. This external attention (EA) mechanism employs learnable memory units that enhance the network's capacity to process complex audio features efficiently. The study demonstrates that EAViT achieves a remarkable overall accuracy of 93.99%, surpassing state-of-the-art models.
Title:
Instruct-DeBERTa: A Hybrid Approach for Aspect-based Sentiment Analysis on Textual Reviews
Authors: Dineth Jayakody, A V A Malkith, Koshila Isuranda, Vishal Thenuwara, Nisansa de Silva, Sachintha Rajith Ponnamperuma, G G N Sandamali, K L K Sudheera
Subjects: Subjects:
Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
Abstract
Aspect-based Sentiment Analysis (ABSA) is a critical task in Natural Language Processing (NLP) that focuses on extracting sentiments related to specific aspects within a text, offering deep insights into customer opinions. Traditional sentiment analysis methods, while useful for determining overall sentiment, often miss the implicit opinions about particular product or service features. This paper presents a comprehensive review of the evolution of ABSA methodologies, from lexicon-based approaches to machine learning and deep learning techniques. We emphasize the recent advancements in Transformer-based models, particularly Bidirectional Encoder Representations from Transformers (BERT) and its variants, which have set new benchmarks in ABSA tasks. We focused on finetuning Llama and Mistral models, building hybrid models using the SetFit framework, and developing our own model by exploiting the strengths of state-of-the-art (SOTA) Transformer-based models for aspect term extraction (ATE) and aspect sentiment classification (ASC). Our hybrid model Instruct - DeBERTa uses SOTA InstructABSA for aspect extraction and DeBERTa-V3-baseabsa-V1 for aspect sentiment classification. We utilize datasets from different domains to evaluate our model's performance. Our experiments indicate that the proposed hybrid model significantly improves the accuracy and reliability of sentiment analysis across all experimented domains. As per our findings, our hybrid model Instruct - DeBERTa is the best-performing model for the joint task of ATE and ASC for both SemEval restaurant 2014 and SemEval laptop 2014 datasets separately. By addressing the limitations of existing methodologies, our approach provides a robust solution for understanding detailed consumer feedback, thus offering valuable insights for businesses aiming to enhance customer satisfaction and product development.
Title:
Multi-Layer Transformers Gradient Can be Approximated in Almost Linear Time
Abstract
The quadratic computational complexity in the self-attention mechanism of popular transformer architectures poses significant challenges for training and inference, particularly in terms of efficiency and memory requirements. Towards addressing these challenges, this paper introduces a novel fast computation method for gradient calculation in multi-layer transformer models. Our approach enables the computation of gradients for the entire multi-layer transformer model in almost linear time $n^{1+o(1)}$, where $n$ is the input sequence length. This breakthrough significantly reduces the computational bottleneck associated with the traditional quadratic time complexity. Our theory holds for any loss function and maintains a bounded approximation error across the entire model. Furthermore, our analysis can hold when the multi-layer transformer model contains many practical sub-modules, such as residual connection, casual mask, and multi-head attention. By improving the efficiency of gradient computation in large language models, we hope that our work will facilitate the more effective training and deployment of long-context language models based on our theoretical results.
Title:
MCTR: Multi Camera Tracking Transformer
Authors: Alexandru Niculescu-Mizil, Deep Patel, Iain Melvin
Subjects: Subjects:
Computer Vision and Pattern Recognition (cs.CV)
Abstract
Multi-camera tracking plays a pivotal role in various real-world applications. While end-to-end methods have gained significant interest in single-camera tracking, multi-camera tracking remains predominantly reliant on heuristic techniques. In response to this gap, this paper introduces Multi-Camera Tracking tRansformer (MCTR), a novel end-to-end approach tailored for multi-object detection and tracking across multiple cameras with overlapping fields of view. MCTR leverages end-to-end detectors like DEtector TRansformer (DETR) to produce detections and detection embeddings independently for each camera view. The framework maintains set of track embeddings that encaplusate global information about the tracked objects, and updates them at every frame by integrating the local information from the view-specific detection embeddings. The track embeddings are probabilistically associated with detections in every camera view and frame to generate consistent object tracks. The soft probabilistic association facilitates the design of differentiable losses that enable end-to-end training of the entire system. To validate our approach, we conduct experiments on MMPTrack and AI City Challenge, two recently introduced large-scale multi-camera multi-object tracking datasets.
Keyword: human object interaction
There is no result
Keyword: visual relation detection
There is no result
Keyword: object detection
Title:
Title:
Title:
Title:
Title:
Keyword: transformer
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Keyword: scene understanding
There is no result
Keyword: visual reasoning
There is no result